ABSTRACT Our collaborative team recently developed a method termed Transition-Path Theory/Markovian Milestoning (TPT/MM) for the prediction of generalized transition rates in all-atom simulations. Its utility was demonstrated in the context of small gas molecule entry and exit in proteins [Yu et al., J Amer Chem Soc 2015;147:3041]. The overarching objective of this project is to develop the TPT/MM approach to handle the binding and unbinding of polyatomic ligands from proteins which involve (a) binding site desolvation and (b) large-scale conformational changes of the protein. Additionally we aim to test new approaches to making TPT/MM even more ef?ecient. The project involves two postdoctoral researchers and two PhD students, one of each at each of the two collaborating institutions (Drexel U. and NYU). Activities will involve code development and large-scale molecular simulations to implement the required biased sampling of (1) binding-site hydration and (2) protein and ligand conformational changes. Test-bed systems of interest include ATP and substrate binding to epidermal growth factor receptor kinase (EGFRK), and substrate and inibitor binding to HIV-1 protease. The ultimate goal is a robust, validated method for estimating ligand on- and off-rates based on 3D all-atom structures available in public databases.